zldrobit_darknet2onnx2tf2tflite2op_tflite.sh

# Note: The author has kept the weights and cfg files strictly in their respective directories.
# If you add yours, do place them in the same folders

!pip install onnx==1.6.0 onnx-tf==1.5.0
!pip install tensorflow-gpu==1.15.0

%cd repo
!wget "https://pjreddie.com/media/files/yolov3-tiny.weights" -O "/weights/" 
# The cfg file is already present in the repo.

# Convert DarkNet -> ONNX
!python3 detect.py --cfg /content/repo/cfg/yolov3-tiny.cfg --weights /content/repo/weights/yolov3-tiny.weights
# The output ONNX file is weights/export.onnx

# Convert ONNX model to TensorFlow (.pb) model
!python3 onnx2tf.py
# The output file is weights/yolov3-tiny.pb

# Convert TensorFlow to TFLite
## optional: Preprocess pb file to avoid NCHW conv, 5-D ops, and Int64 ops
!python3 prep.py
# Proceed to next section to convert to TFLite